WORKING OF THE OPTIONAL REFERENDUM ON LEGISLATIVE ENACTMENTS

The optional referendum upon legislative enactments, like the compulsory referendum upon proposed constitutional amendments, can best be studied by inspection of the results of the votes actually cast upon the measures thus referred to the electorates.

In the first place, popular interest in measures referred to the voters under the optional referendum is more general than in measures referred under the compulsory constitutional referendum. Whilst during the years 1899-1908 only one-tenth of the latter were voted on by three-fourths of those who went to the polls, nearly one-half of the former have been voted on by a similar proportion of the voters. Whilst the majority of the constitutional amendments fail to arouse any perceptible interest in above fifty per cent of the voters, not more than one-fifth of the measures referred under the optional referendum fail to interest a majority of the voters. In short, the optional referendum places upon the voters no such burden of deciding measures which do not interest them as is placed upon them by the compulsory constitutional referendum. At the same time the voters reject a larger proportion of the measures referred to them under the optional referendum than of those referred under the compulsory constitutional referendum. Whilst only about thirty per cent of the latter were rejected, more than half of the measures referred under the optional referendum have been rejected by the voters. In the twelve states in which the optional referendum had been put to use down to the close of 1914 the popular veto was invoked in the cases of seventy-eight measures and was actually exercised in the cases of forty-one. In other words, under the optional referendum the voters are much less frequently required to vote upon measures of which they do not disapprove than under the compulsory referendum. Thus, if the test of the popular referendum is its efficiency as a device for preventing the adoption of measures not satisfactory to the voters, the optional referendum upon legislative enactments is apparently a more efficient instrument than the compulsory referendum upon proposed constitutional amendments.

The greater efficiency of the optional referendum is indicated also by an examination of the nature of the measures referred to the electorates by petition. It has been shown that under the compulsory constitutional referendum many trivial matters can be disposed of only by reference to the voters, and that this condition will persist so long as the constitutions of many states are loaded with legislative detail. Under the optional referendum presumably no measure is referred to the voters unless a substantial number of voters is sufficiently interested to file a petition. In practice it appears that the measures referred by petition are rarely of trivial character. The titles of the fifty legislative measures referred to the electorates by petition in the states possessing the direct popular referendum down to the close of 1912 have been listed by various writers,1 and are readily accessible. Of the twenty-eight legislative measures referred to the state electorates by petition in 1913 and 1914, six related to matters of local government, five to the regulation of public utilities, four to the conduct of higher education, three to compensation for industrial accidents, and the other ten to various topics ranging from the establishment of a penal code to the inspection of investment securities. There is no measure of such trivial character as often appears under the compulsory constitutional referendum.

Examination of the measures vetoed by the state electorates leads to the conviction that the legislatures enacting them were out of touch with, if not in direct opposition to, public opinion upon the matters concerned. Of the twenty-five measures vetoed by the voters, down to 1912, a half-a-dozen were acts providing increased appropriations for public institutions or increased salaries or perquisites for public officials, two provided for the creation of new state offices, and three related to the tenure or compensation of local officers. Two others related to changes in fiscal machinery or procedure. The other vetoed measures related to a variety of subjects, such as the establishment of a summer school for teachers, and the regulation of waterrights for irrigation in Colorado, the organization and use of the state militia in Montana and South Dakota, the apportionment of congressional districts in the latter state, and the standardiza

1 See A. L. Lowell, Public Opinion and Popular Government, app. B. See also Equity, xv, no. 1. (January, 1913.)

tion of the percentage of alcohol permissible in intoxicating liquors in Maine. Among the sixteen measures vetoed by the electorates in 1913 and 1914 were proposals to license prize fighting, restrict the sale of game, require railroads to employ full crews of trainmen, license commission-merchants, and sterilize habitual criminals. There is a corresponding variety in the laws which upon reference to the voters were approved by them. South Dakota and Oregon are the states in which the optional referendum has been most freely used. In the former, the popular veto has been invoked against thirteen measures, in six cases with sucThe vote upon referred measures varied from 86.9 per cent upon the divorce act of 1908, which was sustained, to 60.1 per cent in 1910 upon an amendment to the law relating to a "city, town, or place desiring to become a candidate for county seat." This was also sustained. In Oregon six of the fifteen measures referred to the voters have been vetoed, and the range of public interest in referred measures has been about the same as in South Dakota.

cess.

The evidence with respect to the voting upon measures referred to the people under the optional referendum tends to corroborate the inferences drawn with respect to the voting upon measures referred under the compulsory constitutional referendum. The evidence is perhaps insufficient to warrant any final conclusions, but so far as it goes, it indicates that with respect to the measures in which the voters are interested they are able to vote with discrimination. Since the voters are more generally interested in measures referred under the optional referendum than in those submitted under the compulsory constitutional referendum, it follows that on the whole the operation of the optional referendum is more satisfactory. There is indeed no logical reason why the voters should not be as capable of deciding the fate of statutes referred to them upon petition of a certain fraction of the electorate as of deciding the fact of constitutional amendments referred to them by mandate of the constitution itself. It is not the nature of the procedure under which the reference is made, but the nature of the measure, that determines the action of the voters. The value of right procedure for the selection of measures for reference to the electorate lies in the desirability of excluding from reference such measures as will not interest the

voter. In the case of a measure referred by petition, there is at least the presumption that some of the voters are deeply interested in the measure. In the case of a measure referred by constitutional mandate, there is no such presumption. If the contents of the state constitutions had been restricted to the important matters which alone were originally inserted therein, there would have been such a presumption, but that has not been the case. Under the conditions that prevail in most of the states, the presumption is rather that the bulk of the proposed constitutional amendments will be such as the voters would be glad to leave to the discretion of their representatives if they could. There is, however, no way of doing this, except by abolishing the compulsory constitutional referendum, and extending the optional referendum to all constitutional amendments proposed by the state legislatures.

The substitution of the optional for the compulsory referendum on constitutional amendments is a change in the political system of the states that might well receive consideration. If the legislatures could adopt uncontested amendments without reference to the voters, they would be able to reduce the time required for amendments in such cases by from one to four years. There might often be a great gain to the public in such a saving of time, to say nothing of the advantage of relieving the electorate from a needless burden. The chief objections to the change spring from practical rather than theoretical considerations.

EMERGENCY LEGISLATION

First, there is the difficulty under the optional referendum of giving to the legislature adequate power for dealing with emergencies without imposing undue restrictions upon the power of popular veto. An emergency exists whenever an important public interest demands immediate action, but immediate action may be prevented, if the necessary legislation is opposed by a number of voters sufficient to file a petition for a referendum. Either the power to refer a legislative enactment to the electorate, that is, to suspend the operation of a legislative enactment pending its approval or disapproval by the electorate, must be restricted, or the ability of the legislature to deal

promptly and effectively with emergencies must be seriously impaired.

The states which have adopted the optional popular referendum may be divided into three classes with respect to the manner in which they have met this difficulty. The first class comprises those states which have chosen the first horn of the dilemma.1 In these states the referendum is not applicable to measures enacted for the purpose of dealing with an emergency. Thus in South Dakota the referendum is not applicable to "such laws as may be necessary for the immediate preservation of the public peace, health, or safety, support of the state government and its existing public institutions." In Oregon, however, an emergency may be constitutionally declared only in the case of laws "necessary for the immediate preservation of the public peace, health, or safety." All other measures, including bills making appropriations, except when passed in emergencies as above described, are subject to suspension when referendum petitions are filed against them. In each of the states of this class an emergency is defined in the terms of either the South Dakota or the Oregon constitution. The legislature itself, however, is the judge of the necessity of legislation in cases of alleged emergency, and may declare the existence of the emergency by an ordinary majority vote. If a declaration of emergency is contained in the preamble of a bill, the enactment of the bill serves itself as a declaration of emergency. Therefore, unless the legislatures can be constrained by the courts to give a very strict interpretation to the expressions "public peace," "health," and "safety," which is undesirable even if possible, this solution of the problem imposes extensive restrictions upon the power of popular veto.

The objections to such a solution of the problem are apparent. First, the power of a majority of the legislature to declare an emergency is liable to abuse, for the legislators are under strong temptation to declare an emergency whenever they have reason to fear that they have acted without the sanction of public opinion. Certain legislatures seem almost to have formed the habit of forestalling the exercise of the popular veto as much as possible by passing all measures under color of an alleged emer

1 South Dakota, Oregon, Montana, Oklahoma, Missouri, Arkansas, Colorado, New Mexico, and Washington.